{
"slug": "section-5-ftc-act-autonomous-ai-agents-avoiding",
"topic_id": "TOPIC-058",
"cluster": "Regulatory & Sovereign AI",
"tier": "Tier B",
"title": "Consumer protection liability when the bot makes the promise",
"summary": "Consumer protection law holds a business responsible for representations made on its behalf. Nothing in that framework requires the representation to have come from a person.",
"lede": "A company deploys an agent that negotiates with customers. It offers a refund the company does not honour. The company's position is that the agent made an error, and consumer protection law is unlikely to find that interesting.",
"date": "2024-04-06",
"category": "Future of Work",
"author_id": "constance-ibe-whitmore",
"tags": [
"FTC",
"consumer protection",
"deceptive practices",
"AI agents",
"liability",
"disclosure"
],
"image_title": "When The Bot Makes The Promise",
"schema": "Article",
"key_takeaways": [
"Consumer protection liability attaches to the business for representations made in its name, regardless of what produced them.",
"'The model hallucinated' is an explanation, not a defence — the business chose to deploy a system that makes representations.",
"Bounding what an agent can commit to, and giving the consumer a record of it, addresses both the liability and the dispute."
],
"body": [
{
"type": "h2",
"text": "The framework, briefly"
},
{
"type": "diagram",
"kind": "flow",
"alt": "The agent proposes, the policy engine decides, and the consumer's record comes from the engine",
"caption": "If the model both makes the promise and writes the record of it, a dispute has no independent evidence on either side.",
"nodes": [
{
"label": "Agent proposes a remedy",
"note": "generated"
},
{
"label": "Policy engine checks permitted types and value",
"note": "the control",
"good": true
},
{
"label": "Remedy rendered from policy",
"note": "not from the model",
"good": true
},
{
"label": "Consumer receives a verifiable record",
"note": "independent of the model",
"good": true
}
]
},
{
"type": "p",
"html": "Consumer protection regimes generally prohibit deceptive acts or practices. A representation is typically deceptive if it is likely to mislead a reasonable consumer and is material to their decision."
},
{
"type": "p",
"html": "Two features matter here. Intent is generally not required. And the business bears responsibility for representations made in its name through its channels."
},
{
"type": "h2",
"text": "Why the usual defences do not fit"
},
{
"type": "table",
"head": [
"Argument",
"Problem"
],
"rows": [
[
"The model produced unintended output",
"The business chose to deploy a system capable of making representations"
],
[
"The consumer should have known it was a bot",
"The representation was still made in the business's name"
],
[
"Terms of service disclaim agent statements",
"Disclaimers rarely cure a deceptive representation that induced the transaction"
],
[
"It was a one-off error",
"Systems produce errors at a rate; a rate is a practice"
]
]
},
{
"type": "p",
"html": "The last row is the one businesses underestimate. An individual employee's misstatement is an incident. A deployed system's error rate across thousands of interactions looks structural, and structural is how these cases are framed."
},
{
"type": "h2",
"text": "Four representation types worth bounding"
},
{
"type": "ol",
"items": [
"Price and terms. Anything about cost, duration, renewal or cancellation.",
"Remedies. Refunds, credits, replacements, waivers of fees.",
"Capability. What the product or service does, particularly in regulated categories.",
"Eligibility. Whether a consumer qualifies for something."
]
},
{
"type": "p",
"html": "These are the classes where a representation is both material and easy for a generative system to produce plausibly. They are also enumerable, which makes them boundable."
},
{
"type": "h2",
"text": "Bounding at the boundary, not the prompt"
},
{
"type": "p",
"html": "Instructing a model not to promise refunds is a request. Preventing it from issuing one is a control."
},
{
"type": "code",
"text": "# The agent proposes; the boundary decides\n\ndef issue_remedy(agent_request, conversation):\n policy = remedy_policy_for(conversation.customer, conversation.issue)\n\n if agent_request.type not in policy.permitted_types:\n return refuse(\"outside permitted remedies\")\n if agent_request.value > policy.max_value:\n return escalate_to_human(agent_request, conversation)\n\n remedy = policy.render(agent_request) # canonical, from policy\n receipt = sign_issuer(remedy) # business attests\n deliver_to_consumer(remedy, receipt)\n return remedy\n\n# The consumer's record comes from the policy engine,\n# not from the model's description of what it offered."
},
{
"type": "p",
"html": "The last comment is the substantive design point. If the consumer's record of what was promised is generated by the same model that made the promise, a dispute about what was said has no independent evidence on either side."
},
{
"type": "h2",
"text": "What the consumer should receive"
},
{
"type": "p",
"html": "A record that is useful to them, which incidentally is also the record that protects the business."
},
{
"type": "ul",
"items": [
"What was agreed, in specific terms — amount, timing, conditions",
"When, and through which channel",
"That it was an automated interaction, where that is not obvious",
"Something verifiable, so the record is not simply the business's assertion"
]
},
{
"type": "p",
"html": "A consumer holding this is less likely to complain to a regulator, because they have what they need. A business holding the counterpart is in a far better position if the consumer does."
},
{
"type": "h2",
"text": "The disclosure question"
},
{
"type": "p",
"html": "Whether a consumer must be told they are interacting with an automated system is an area of active regulatory attention, and requirements differ by jurisdiction and context."
},
{
"type": "p",
"html": "The practical position is that disclosure is cheap and its absence is an easy criticism. A business that discloses clearly and bounds what the agent can commit to has addressed both the disclosure question and the substance, without waiting for the rule to settle."
},
{
"type": "p",
"html": "This describes consumer protection principles at a general level and is not legal advice. Standards, enforcement posture and disclosure requirements vary by jurisdiction; take specifics to counsel."
},
{
"type": "h2",
"text": "Four representation classes worth bounding"
},
{
"type": "table",
"caption": "Where a generative system can produce a material misstatement",
"head": [
"Class",
"Examples"
],
"rows": [
[
"Price and terms",
"Cost, duration, renewal, cancellation"
],
[
"Remedies",
"Refunds, credits, replacements, fee waivers"
],
[
"Capability",
"What the product does, especially in regulated categories"
],
[
"Eligibility",
"Whether a consumer qualifies for something"
]
]
},
{
"type": "p",
"html": "These are enumerable, which is what makes them boundable. Instructing a model not to promise refunds is a request; preventing it from issuing one is a control."
},
{
"type": "h2",
"text": "Objections and honest limits"
},
{
"type": "p",
"html": "“Our terms of service disclaim agent statements.” Disclaimers rarely cure a deceptive representation that induced a transaction. The statement was made in the business's name through the business's channel."
},
{
"type": "p",
"html": "“It was a one-off error.” Systems produce errors at a rate, and a rate is a practice. An individual employee's misstatement is an incident; a deployed system's error rate across thousands of interactions reads as structural, and structural is how these cases are framed."
},
{
"type": "p",
"html": "Disclosure is an area of active regulatory attention and it is cheap. A business that discloses clearly and bounds what the agent can commit to has addressed both the disclosure question and the substance without waiting for the rule to settle."
}
],
"faq": [
{
"q": "Can a business disclaim its agent's statements?",
"a": "Terms of service disclaimers rarely cure a deceptive representation that induced a transaction. The representation was made in the business's name."
},
{
"q": "Does intent matter?",
"a": "Generally not under deception standards. A representation likely to mislead a reasonable consumer, material to their decision, is the test."
},
{
"q": "How should agent commitments be bounded?",
"a": "At the system boundary rather than in the prompt. A policy engine decides what remedies are permitted; the model proposes and does not issue."
},
{
"q": "Why shouldn't the model write the consumer's record?",
"a": "Then a dispute about what was promised has no independent evidence. The record should come from the policy engine that issued the remedy."
}
],
"sources": [
{
"t": "FTC Policy Statement on Deception",
"u": "https://www.ftc.gov/legal-library/browse/ftc-policy-statement-deception"
},
{
"t": "State unfair and deceptive acts and practices statutes."
},
{
"t": "Regulatory guidance on automated interaction disclosure."
},
{
"t": "Published enforcement actions involving automated customer interactions."
},
{
"t": "Federal Trade Commission Act, Section 5",
"u": "https://www.ftc.gov/legal-library/browse/statutes/federal-trade-commission-act"
}
],
"related": [
{
"slug": "my-agent-did-defense-businesses-will-dispute",
"title": "The 'my agent did it' defence",
"category": "Future of Work"
},
{
"slug": "consumer-mandate-receipt",
"title": "Visa made cardholders responsible for their agents",
"category": "Vertical"
},
{
"slug": "ai-agent-act-senator-warners-draft-legislation",
"title": "Reading draft agent legislation",
"category": "Future of Work"
}
],
"image": "https://cdn.twc.sh/images/igcache/When%20The%20Bot%20Makes%20The%20Promise/1500_900/blog.jpg",
"wordcount": 875,
"url": "/blog/section-5-ftc-act-autonomous-ai-agents-avoiding.html",
"reading_time": "4 min read",
"meta_description": "Consumer protection law holds a business responsible for representations made in its name. Nothing requires those to come from a person.",
"hub": {
"slug": "topics/ai-oversight-regulation",
"title": "AI oversight and regulation"
},
"answer": "Very likely. Section 5 of the FTC Act reaches representations likely to mislead a reasonable consumer and material to their decision. Intent is generally not required, and nothing in the framework asks what produced the representation. “The model hallucinated” is an explanation, not a defence.",
"answer_q": "Is a business liable when its AI agent makes a promise it will not honour?",
"entities": [
{
"name": "Federal Trade Commission",
"type": "GovernmentOrganization",
"url": "https://www.ftc.gov/legal-library/browse/statutes/federal-trade-commission-act",
"primary": true
}
],
"glossary": [
{
"term": "Deception",
"def": "A representation likely to mislead a reasonable consumer, material to their decision. Intent is generally not an element."
},
{
"term": "Material representation",
"def": "One likely to affect a consumer's choice. Price, remedies and eligibility almost always qualify."
},
{
"term": "Policy engine",
"def": "A component that decides what the business is willing to offer, independent of what a model proposes."
}
],
"checklist": {
"title": "Bounding what an agent can commit to",
"id": "bound",
"desc": "Four steps at the system boundary.",
"steps": [
{
"name": "Enumerate the representation classes.",
"text": "Price, remedies, capability, eligibility. Four lists, each short."
},
{
"name": "Move the decision to a policy engine.",
"text": "The model proposes; the engine decides what is permitted and at what value."
},
{
"name": "Render the consumer's record from the engine.",
"text": "Not from the model's description of what it offered."
},
{
"name": "Give the consumer something verifiable.",
"text": "A record they can check is one they are less likely to complain about, and one that protects you if they do."
}
]
},
"cta": {
"title": "Where this fits in Manav",
"html": "Manav's role here is the consumer-side record: the remedy is rendered from the policy engine, countersigned by the business, and delivered to the consumer in a form either party can verify later without the other's cooperation.",
"href": "../docs.html",
"label": "See consumer receipts"
}
}